Effect of combination therapy with alogliptin and ...

Endocrine Journal 2014, 61 (10), 1031-1039

GASTRIN is a hormone secreted from G cells in the gastric pyloric gland that regulates gastric acid secre-tion in the parietal cells of the stomach. Gastrin also promotes β-cell neogenesis from pancreatic duct cells and increases islet mass in a rat-model with ligated pancreatic ducts [1]. In humans, Meier et al. recently reported a case showing increased pancreatic β-cell replication adjacent to gastrinomas, suggesting a direct effect of gastrin on pancreatic tissues [2]. Proton pump inhibitors (PPI), which are widely used clini-cally for the treatment of gastro-esophageal reflex dis-eases, gastritis due to excess stomach acid, and gastric

Effect of combination therapy with alogliptin and lansoprazole on glycemic control in patients with type 2 diabetes

Kohzo Takebayashi1), Shintaro Sakurai1), Tatsuhiko Suzuki1), Kenichiro Hori1), Tomoko Terasawa1), Rika Naruse1), Kenji Hara1), Mariko Suetsugu1), Takafumi Tsuchiya1), Hiromi Aoki2), Takashi Hamasaki3), Hiroshi Shuutou4) and Toshihiko Inukai1)

1) Department of Internal Medicine, Dokkyo Medical University Koshigaya Hospital, Koshigaya, Japan2) Higashiwashinomiya Hospital, Kuki, Japan3) Hamasaki Clinic, Kasukabe, Japan4) Iryouhoujinkenshinkai-minamikoshigayakenshinkai Clinic, Koshigaya, Japan

Abstract. The main purpose of the current study was to investigate the effect of a combination of alogliptin [a dipeptydil peptidase (DPP)-4 inhibitor] and lansoprazole [a proton pump inhibitor (PPI)] compared with alogliptin mono-therapy on glycemic control in patients with type 2 diabetes. This study was a multicenter randomized open-label study. One hundred type 2 diabetic patients were randomly assigned to either the alogliptin with lansoprazole group or the alogliptin mono-therapy group. After 3 months of treatment, the changes in hemoglobin (Hb)A1c, fasting plasma glucose (FPG), serum gastrin, homeostasis model assessment (HOMA)-β, and HOMA-insulin resistance (IR) were evaluated. A significant decrease in HbA1c and FPG, and a significant increase in HOMA-β were observed in both groups (all with P <0.0001). However, there were no significant differences in changes in HbA1c, FPG, or HOMA-β before and after therapy between the combination and alogliptin mono-therapy group (P =0.2945, P =0.1901, P =0.3042, respectively). There was a significant elevation of serum gastrin in the combination group compared with the alogliptin mono-therapy group (P =0.0004). This study showed that, although combination therapy with alogliptin and lansoprazole more effectively elevated serum gastrin levels compared with alogliptin mono-therapy, the effect of the combination therapy on glycemic control was equal to that of alogliptin mono-therapy during a 3-month study period.

Key words: Gastrin, DPP4 inhibitor, Type 2 diabetes

ulcers, can indirectly elevate serum gastrin levels [3-8]. Interestingly, in animal models of type 2 diabetes, it is reported that PPI treatment improves glycemic control [9], probably via its possible effect in increasing serum gastrin levels. Furthermore, in human retrospective or cross-sectional studies, an association between admin-istration of PPI and better glycemic control in patients with type 2 diabetes was reported [10-12]. However, two recent prospective randomized double blind pla-cebo-controlled studies using PPI in a small number of patients with type 2 diabetes showed conflicting results of its effect on glycemic control [13, 14].

Importantly, recent evidence suggests the potentially beneficial effect of combination therapy of various hor-mones rather than single hormone therapy [15]. In fact, combination therapy with glucagon-like peptide-1 (GLP-1) and gastrin more effectively improves hyper-glycemia than single therapy by each hormone in non-

Original

©The Japan Endocrine Society

Submitted May 10, 2014; Accepted Jul. 16, 2014 as EJ14-0208Released online in J-STAGE as advance publication Sep. 2, 2014Correspondence to: Kohzo Takebayashi, M.D., Department of Internal Medicine, Dokkyo Medical University Koshigaya Hospital, 2-1-50, Minami-Koshigaya, Koshigaya 343-8555, Japan.E-mail: [email protected]

1032 Takebayashi et al.

use of a placebo and a double-blind design are not usu-ally permitted as a post-marketing clinical study in Japan. 100 patients were recruited from December 2012 to August 2013 in 5 institutions (2 hospital includ-ing Dokkyo Medical University Koshigaya Hospital and 3 medical clinics) for this study named APPLE; study of combination effect of AlogliPtin and lansoPra-zoLE on glycemic control in patients with type 2 diabe-tes. This study was registered at UMIN 0000009445. After enrollment, patients were assigned to two groups [alogliptin alone (n =50) and alogliptin + lansopra-zole (n =50)], by a central computer system based on HbA1c, BMI, gender, and age.

Key inclusion criteria were 1) patients aged 20 or over, 2) patients diagnosed as type 2 diabetes according to the Japan Diabetic Society Criteria [21], 3) outpatient status, 4) patients with 6.5% and greater HbA1c levels, and with 0.5% or less variation of HbA1c in 1 month prior to the study, 5) patients who did not receive DPP4 inhibitors during 1 month prior to the study, 6) patients in whom type and dose of all drugs including anti-dia-betic ones was not changed during 1 month prior to the study, and 7) patients giving consent in writing for par-ticipation for this study. Key exclusion criteria were 1) patients with severe ketosis, diabetic coma, or type 1 diabetes, 2) patients with severe infectious disease, within 1 month after surgery, or having a severe trau-matic injury, 3) patients with allergy to either alogliptin or lansoprazole, 4) patients under atazanavir-treatment, 5) patients treated with PPIs of all types or H2 block-ers of all types, 6) patients treated with insulin, and 7) patients otherwise judged as unacceptable for partici-pation in this study by our medical doctor.

Clinical and laboratory tests showed no evidence of liver dysfunction or autoimmune disease in any patients. No patients showed serum creatinine levels exceeding 1.1 mg/dL. Patients visited the hospital and had blood and urine tests basically monthly. Seven patients in the alogliptin alone arm dropped out of the study. One patient discontinued the drug due to epi-gastric pain, and 1 patient was released due to suspi-cion of pancreatic cancer. Four patients were found to have poor compliance with the drugs. One patient was found to have had HbA1c <6.5% on enrollment. Four patients dropped from the study in the combi-nation group with alogliptin and lansoprazole. One patient withdrew due to a mild cerebral infarction, and 1 patient withdrew because of implantation of a pace-maker. One patient noticed occasional hypoglycemic

obese diabetic (NOD) mice (animal model of type 1 diabetes) [16]. This result is also supported in the same animal model by combination therapy with dipeptidyl peptidase (DPP)-4 inhibitors, which block degrada-tion of GLP-1 by DPP4 resulting in the elevation of serum active GLP-1 levels [17], and PPI. Although these studies were performed in type 1 diabetic animal models, a very recent study showed that combination therapy with exendin-4 (a GLP-1 receptor agonist) and omeprazole (a PPI) more effectively improved glyce-mic control than single therapy for these drugs in db/db mice (an animal model of obesity and type 2 diabe-tes) [18]. Furthermore, in db/db mice, a GLP-1-gastrin dual agonist has showed a more sustained gluco-reg-ulatory effect accompanied with significant increase of β-cell mass in pancreatic tissue than liraglutide (a GLP-1 receptor agonist) monotherapy [19]. Taken together, these studies suggest the possibility that, also in type 2 diabetes patients, combination therapy with DPP4-inhibitors and PPI may provide a more benefi-cial effect for glycemic control than mono-therapy with DPP4 inhibitors. However, there is no evidence show-ing the combination-effect of DPP4-inhibitors and PPI for glycemic control in patients with type 2 diabetes. Based on these findings, we investigated the combina-tion effect of alogliptin (a DPP4 inhibitor) and lanso-prazole (a PPI) on glycemic control, compared with the effect of alogliptin alone in patients with type 2 diabe-tes in this multicenter, randomized study. We hypothe-sized that combination therapy with alogliptin and lan-soprazole would provide better glycemic control than alogliptin mono-therapy.

Patients and Methods

PatientsThe sample size was determined based on the effect

of alogliptin on HbA1c levels because we predicted that combination-therapy can reduce HbA1c levels at least to the same degree as that observed with alo-gliptin mono-therapy. Based on the previous study per-formed in Japan [20], we hypothesized that alogliptin can achieve an 0.8% reduction of HbA1c, and based on our clinical experience in our hospital, a mean ± stan-dard deviation (SD) of HbA1c at baseline was sought to be 8.0 ± 1.0%. With a power of 90% and α of 0.05 two tailed, a sample size of 36 per group would be required.

This study was a multicenter randomized open-label study but it was not a double-blind study, because the

1033Combination therapy with alogliptin and lansoprazole

which PPI significantly decreased the HbA1c levels in 3 months study period.

Primary outcomes of this study for each group were HbA1c and fasting plasma glucose (FPG) levels 3 months after starting the study. Key secondary out-comes for each group were fasting serum gastrin and plasma insulin levels, HOMA-IR, HOMA-β, and body weight 3 months after starting the study. No change in drug administration was made for any patient during the 3 months after starting the study.

Blood and urine were basically collected in the outpa-tient department from 8:30 to 9:30 a.m. after overnight fasting for at least 10 hr. After the blood was collected, it was immediately placed into specific test tubes for different assays, and then rapidly centrifuged at 1,500 rpm for 5 minutes to separate the serum or plasma from the clot-containing blood cells. Except for samples for glucose, HbA1c, and lipid measurements, the samples were stored frozen at -70°C until analysis. At the time of sampling, the patients were weighed. Measurement of plasma glucose, hemoglobin (Hb) A1c, and serum lipids

Fasting plasma glucose (FPG) was evaluated imme-diately after blood collection using an automated glu-cose oxidase method (Glucose Auto Stat GA1160®;

symptoms after starting the drugs. One patient was found to have poor compliance with the drugs. As a result, 43 patients in the alogliptin mono-therapy group and 46 in the combination group with alogliptin and lansoprazole completed the study.

The characteristics of the patients treated with alo-gliptin alone and with alogliptin and lansoprazole are shown in Table 1.

MethodsThe use of lansoprazole is non-approved indication

for the treatment of diabetes in Japan. Therefore lanso-prazole was administrated to the patients without insur-ance treatment in Japan; we covered the related cost of lansoprazole administration instead of patients. There would be an ethical problem because of the non-ap-proved use of this drug if the unexpected side effect occurred by long-term administration of this drug. Thus, we decided to choose the relatively short-term study period (3 months) although it appears to be more effective in longer term treatment as a year if the mech-anism on possible glucose-lowering effect of PPI is based on the improvement of β-cell function. Another reason why we chose 3 months as study period was based on the results of the study in Singh et al. [13] in

Table 1 Clinical characteristics of the patients treated with alogliptin alone and alogliptin and lansoprazole Alogliptin Alogliptin + lansoprazole P

No. (male/female) 43 (23/20) 46 (24/22) -Age (year) 65.8 ± 12.4 64.8 ± 9.2 0.6554BMI (kg/m2) 24.9 ± 4.3 25.4 ± 4.1 0.6412FPG (mg/dL) 153.6 ± 34.4 159.7 ± 34.5 0.4065HbA1c (%) 7.7 ± 0.5 7.6 ± 0.6 0.5205Diabetic therapy

SU (S1/S2/S3/S4/S5/S6/S7/S8) 6/6/7/2/1/3/2/0 8/7/9/0/1/1/2/1 -MET(M1/M2/M3/M4) 2/1/3/1 3/0/1/1 -αGI alone 2 2 -Pioglitazone alone 1 1 -

Antihypertensive drugs(A/C/AC) 3/2/10 6/2/12 -

Statins 15 13 -Data are expressed as mean ± standard deviation (SD), Comparison in variables between two groups were made by use of an unpaired t test. P: p value, P <0.05 are defined as statistical significance. BMI, body mass index; FPG, fasting plasma glucose; Diabetic therapy, the number of the patients with respective diabetic therapies; SU, sulfonylurea; MET, metformin; α-GI, α glucosidase inhibitor; S1:SU alone, S2: SU and MET, S3: SU, MET and α-GI, S4: SU, MET, α-GI and pioglitazone, S5: SU, MET and pioglitazone, S6: SU and pioglitazone, S7: SU and α-GI, S8: SU, pioglitazone and α-GI, M1: MET alone, M2: MET and pioglitazone, M3: MET and α-GI, M4: MET, pioglitazone and α-GI. Antihypertensive drugs: the number of the patients with respective antihypertensive drugs, A: angiotensin-II receptor blocker (ARB), C: calcium channel blocker (CA), AC: ARB and CA


or a Welch t-test was chosen based on the homogene-ity of variance calculated by an F-test. Comparisons of various markers in alogliptin- and those in alogloptin with lansoprazole-treated patients for changes from baseline were assessed using an analysis of variance (ANCOVA) model with treatment as fixed effects and the corresponding baseline value as a covariate. For HbA1c and FPG, least square (LS) mean difference were estimated for the comparison of alogliptin and alogliptin with lansoprazole. Comparison of the pro-portion of patients who achieved HbA1c equal to or less than 6.5% and 7.0 % in two groups was analyzed by the χ2 test. In the ANCOVA and single regression analysis, values of gastrin were log10-transformed. A P value of less than 0.05 was accepted as indicating sta-tistical significance (two-sided).

Results

Changes in various variables at baseline and at 3 months in patient-groups treated with either alogliptin alone or alogliptin and lansoprazole are presented in Table 2. A significant decrease was obtained in HbA1c and in FPG in both groups. However, there was no significant difference in the change in HbA1c and in FPG before and after therapy based on analysis using ANCOVA between the two groups; LS mean changes with standard error (SE) of HbA1c at 1, 2 and 3 months were respectively 0.567 ± 0.158, 0.927 ± 0.185 and 0.997 ± 0.086% in the alogliptin alone group, and those were respectively 0.651 ± 0.158, 1.000 ± 0.185 and 0.839 ± 0.083% in the combination group. LS mean change with SE of FPG at 1, 2 and 3 months were respectively 16.0 ± 4.3, 15.0 ± 4.1, 26.8 ± 3.4 mg/dL in the alogliptin group, and those were respectively 9.3 ± 4.5, 15.3 ± 4.2, 21.6 ± 3.2 mg/dL in the combination group (Fig. 1A, B). In subgroups of patients with sulfonylurea (± other anti-diabetic drugs), with metformin (± other anti-diabetic drugs), and without anti-diabetic drugs at baseline in both treatment groups, there were no significant differ-ence in the change in HbA1c and in FPG at baseline and at 3 months after therapy between two treatment groups (P =0.0945, 0.2978, 0.2146 in HbA1c, and P =0.4361, 0.4060, 0.3888 in FPG, respectively).

The proportion of patients who achieved an HbA1c equal to or less than 6.5% and 7.0 % was respectively 35% and 84% in the alogliptin alone group, and 43% and 72% in the combination group (no significant dif-ference between the two groups). For serum gastrin

Arkray, Kyoto, Japan). HbA1c was also measured immediately after collection of blood in a test tube containing EDTA-2K using high-performance lipid chromatography (HPLC; Hi-Auto A1C®, HA8150®; Arkray). Only HbA1c is detected with this method and the normal range is 4.5% to 6.2% [NGSP: national glycohemoglobin standardization program]. Serum total cholesterol (TC), low-density lipoprotein cho-lesterol (LDL-C), high-density lipoprotein choles-terol (HDL-C), and serum triglyceride (TG) were mea-sured using enzymatic assays. The Determiner L TC II® and Determiner L TG® reagents (Kyowa Medics, Tokyo, Japan) were used for measurements of TC and TG, respectively. HDL-C was measured directly by a method based on selective solubilization of different lipoproteins by proprietary detergents using Cholestest N HDL-C® (Daiichi Pure Chemicals, Tokyo, Japan). LDL-C was also measured directly by a homoge-nous method using Cholestest LDL® (Daiichi Pure Chemicals), rather than indirectly using the Friedwald equation. Serum gastrin was measured by a radioim-munoassay (RIA)-PEG method using the Gastrin RIA kit II® (TFB, Tokyo, Japan). The normal value of fast-ing gastrin using this kit was less than 200 pg/mL. Plasma insulin was measured by a chemiluminescent enzyme immunoassay (CLEIA) using the lumipulse Presto Insulin Kit ® (Fujirebio, Tokyo, Japan)Homeostasis model assessment-insulin resistance (HOMA-IR), and Homeostasis model assessment β (HOMA-β)

HOMA-IR was used as an indicator of insulin resis-tance and was calculated as follows: HOMA-IR = FPG (mg/dL) × fasting immunoreactive insulin (μU/mL)/405. HOMA-β was used as an indicator of insu-lin secretion activity and was calculated as follows: HOMA-β =360 × fasting immunoreactive insulin (μU/mL)/fasting plasma glucose (mg/dL)-63. Ethical considerations

All subjects gave informed consent for inclusion in the study. The study was approved by the Local Ethics Committee in our hospital and was performed accord-ing to the guidelines of the Declaration of Helsinki. Statistical methods

The two time points for each parameter for an indi-vidual were compared using a paired t-test, except those for gastrin, for which a Wilcoxon signed rank test was used as a non-parametric test due to its skewed distributions. Comparisons between the two groups were made using an unpaired t-test for normally dis-tributed data (confirmed by a χ2 test); a Student t-test


Table 2 Changes in various variables at baseline and 3 months after therapy

Alogliptin Alogliptin + lansoprazole

Baseline 3 months P (Difference: 95%CI) Baseline 3 months P (Difference: 95%CI) P*

FPG (mg/dL) 153.6 ± 34.4 128.5 ± 26.6 0.0001(17.0-33.2)* 152.0 ± 35.6 127.3 ± 27.4 <0.0001(16.7-32.7)* 0.1901

HbA1c (%) 7.7 ± 0.5 6.7 ± 0.5 <0.0001(0.8-1.2)* 7.6 ± 0.6 6.8 ± 0.7 <0.0001(0.6-1.0)* 0.2945

TC (mg/dL) 184.4 ± 29.5 178.7 ± 28.9 0.1705(-2.6-,13.8) 194.9 ± 32.1 181.0 ± 36.1 0.0083(3.9-24.0)* 0.3439

TG (mg/dL) 120.3 ± 129.1 110.7 ± 102.5 0.1757(-4.4,23.5) 123.4 ± 68.5 141.3 ± 158.8 0.3042(-52.6,16.8) 0.1532

HDL-C (mg/dL) 58.2 ± 14.9 56.9 ± 13.7 0.3154(-1.3,4.0) 53.4 ± 14.5 49.3 ± 14.0 0.0001(2.2-6.1)* 0.0218*

LDL-C (mg/dL) 105.0 ± 24.8 99.4 ± 25.6 0.0580(-0.2,11.3) 108.7 ± 30.6 100.2 ± 26.1 0.0110(2.0-14.9)* 0.7325

SBP (mm Hg) 131.7 ± 10.8 127.9 ± 12.9 0.0253(0.5, 7.1)* 132.9 ± 10.6 128.3 ± 11.1 0.0422(0.2-9.2)* 0.9223

DBP (mm Hg) 74.5 ± 9.9 74.3 ± 10.0 0.8688(-2.6, 3.1) 75.9 ± 9.6 74.3 ± 10.6 0.3691(-2.0,5.2) 0.6779

BW (kg) 64.4 ± 15.0 64.0 ± 15.0 0.0398(0.02, 0.7)* 66.2 ± 12.0 65.6 ± 11.6 0.0369(0.04,1.2)* 0.4427

Insulin (μU/mL) 7.7 ± 5.7 7.4 ± 6.4 0.7776(-1.6, 2.1) 9.8 ± 8.2 10.0 ± 8.6 0.9065(-2.3, -2.0) 0.4635

Gastrin (pg/mL) 110(68, 190) 120(77.5, 265) 0.00085* 97(70, 110) 180(110, 400) <0.0001* 0.0004*

HOMA-IR 3.0 ± 2.2 2.4 ± 2.3 0.0944(-0.1, 1.2) 4.0 ±3.7 3.6 ± 3.5 0.4643(-06, 1.3) 0.7781

HOMA-β 35.3 ± 34.8 48.1 ± 44.9 0.0171(-23.1, -2.4)* 37.6 ± 33.4 52.9 ± 55.6 0.0242(-28.4, -2.1)* 0.3042P: p value for change before and after therapy by paired t test. P*: p value for comparisons in alogliptin and in alogliptin with lansoprazole group for changes from baseline assessing using analysis of variance (ANCOVA) model with treatment as fixed effects and corresponding baseline value as covariate. P <0.05 are defined as statistically significant (*). CI, confidence intervalData are expressed as means ± SD except for gastrin which are expressed as medians with interquartile ranges (25th and 75th percentile). FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL, low-density lipoprotein cholesterol; SBP, systolic blood pressure; DBP, diastolic blood pressure; BW, body weight; HOMA-IR, Homeostasis model assessment insulin resistance; HOMA-β, Homeostasis model assessment β

Fig.1 LS mean difference in HbA1c (A) and FPG (B) before and after 3 months in one group treated with alogliptin and in another group treated with alogliptin and lansoprazole in patients with type 2 diabetes


levels, there was significant elevation before and after therapy in both groups, however, the difference in gas-trin (log10-transformed) before and after therapy as ana-lyzed using ANCOVA showed a significant elevation of gastrin in the combination group when compared with those in the alogliptin alone group (Fig. 2A, B).

A significant decrease in both SBP and BW and a sig-nificant elevation in HOMA-β by treatment was obtained in both groups, and a significant decrease in TC, LDL-C and HDL-C was found in the combination group.

There was no correlation between log10-trans-formed gastrin and HbA1c, FPG, insulin, HOMA-IR, or HOMA-β in baseline in both groups.

The correlation between the difference in log10-transformed gastrin and those in various markers before and after therapy in two groups is presented in Table 3.

There were no apparent side effects in both groups in this study except for hypoglycemic symptom recog-nized in 1 patient in combination group.

Discussion

In the current study, we investigated the effect of combination therapy with alogliptin (a DPP4 inhibi-tor) and lansoprazole (a PPI) on glycemic control in patients with type 2 diabetes and compared the find-ings with that of alogliptin mono-therapy. Contrary to our expectation, the effects on glycemic control of the two therapies were equivalent. The results in this study do not support those found in previous studies in an animal model of type 2 diabetes with either combina-tion-therapy of GLP-1 receptor agonist and PPI [18], or with GLP-1-gastrin dual agonist [19]. It is difficult to explain the reason for the unexpected result in this study. However, the observational period of the ret-rospective study on glycemic control with PPI mono-therapy in patients with type 2 diabetes, which showed positive results on glycemic control, was longer than 1 year [10]. One recent prospective study in patients with type 2 diabetes treated with PPI mono-therapy for

Fig. 2 Change in serum gastrin before and after 3 months in one group treated with alogliptin (A) and in another group treated with alogliptin and lansoprazole (B) in patients with type 2 diabetes

Table 3 Correlation of difference before and after therapy in log-transformed gastrin with those in various markers

Correlation with ΔgastrinAlogliptin group

R PΔinsulin 0.3592 0.0369*

ΔFPG 0.1621 0.3521ΔHbA1c 0.0128 0.9419ΔHOMA-IR 0.4226 0.0179*

ΔHOMA-β 0.3631 0.0447*

Alogliptin + lansoprazole groupR P

Δinsulin 0.2622 0.0925ΔFPG 0.1818 0.2254ΔHbA1c 0.1508 0.3467ΔHOMA-IR 0.2984 0.0650ΔHOMA-β 0.1659 0.3128

Total patients (total of both groups)R P

Δinsulin 0.2668 0.0207*

ΔFPG 0.1554 0.1801ΔHbA1c 0.1443 0.2136ΔHOMA-IR 0.3139 0.0081*

ΔHOMA-β 0.2193 0.0694Δ, the difference in each marker before and after therapy, FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; HOMA-IR, Homeostasis model assessment insulin resistance; HOMA-β, Homeostasis model assessment β


3 months using pantoprazole significantly decreased the HbA1c levels [13], while in another study using esomeprazole for the same observation period [14], there was no effect. Because the possible effect of gastrin on glycemic control appears to be at least par-tially dependent on the effect of gastrin (or gastrin with GLP-1) [12] for replication of pancreatic β cells which would probably take relatively long term, a longer observation-period of more than a year may be needed to find the effect on glycemic control of single PPI or combination therapy of PPI with DPP4.

Interestingly, although there was no correlation in the present study between serum gastrin levels and plasma insulin levels, HOMA-β, or HOMA-IR in base-line in both groups, a significantly positive correlation was observed between differences before and after therapy in serum gastrin levels and in plasma insulin levels, HOMA-β or HOMA-IR in the alogliptin mono-therapy group. Therefore, speculatively, the change in serum gastrin levels might basically correlate with the change in insulin secretion activity as evaluated by HOMA-β or that in insulin resistance as evaluated by HOMA-IR under the condition in which change in gastrin levels is not so large (i.e., within physiologi-cal variation range). However, even under this con-dition, gastrin does not probably fully affect glycemic control, since gastrin might be associated with both insulin secretion and insulin resistance. Notably, these correlations were attenuated in the combination ther-apy group. Therefore, excess elevation of gastrin by drugs, such as PPI, appears not always to be associated with an increase in insulin secretion. This may also be based on the relatively short-term (3 months) observa-tion in this study.

In a previous study using pantoprazole [13], the dose of the drug was approximately twice that used clinically used for gastro-esophageal reflux disease. The dose of lansoprazole in this study (15 mg/day) was what is usually used in treatment for gastilitis or gastric ulcer. However, in the current study, the degree of ele-vation of serum gastrin levels in combination therapy was approximately twice, and was higher than that (1.5 hold) in the study using pantoprazole. Therefore, the dose of lansoprazole in this study appears to have been basically effective for the elevation for serum gastrin. However, treatment longer than 3 months would be expected to further increase serum gastrin levels [22, 23]. It may be possible that, especially in combina-tion therapy with DPP4 inhibitors and PPI unlike PPI

mono-therapy, more highly-elevated serum gastrin lev-els are needed to show the benefit on glycemic control exceeding that of DPP4 inhibitor mono-therapy itself. Therefore, also from this aspect, it may be interesting to observe the combinational effect with DPP4 inhibi-tors and PPI for a longer observation period.

We investigated the combination effect of a DPP4-inhibitor and PPI. In animal studies in which a synergis-tic effect on glycemic control was obtained [18, 19], a GLP-1 receptor agonist with PPI or gastrin-GLP-1 dual agonist instead of DPP4-inhibitors with PPI was used. It is well known that the elevation of serum GLP-1 lev-els with a DPP4-inhibitor is a milder treatment com-pared with that of a GLP-1 receptor agonist. Therefore, even if the degree of elevation of serum gastrin by PPI is high, it may be possible that a GLP-1 receptor agonist with PPI will be more synergistic for glycemic control than DPP4-inhibitors with PPI. Accordingly, it would be interesting in the future to investigate the effect of a combination therapy with PPI using a GLP-1 receptor agonist on glycemic control.

Recently, Wallace JL et al. reported that PPI can induce dysbiosis [24]. It is known that dysbiosis is con-nected with metabolic syndrome. Therefore, specula-tively, PPI may worsen glycemic control in this aspect, and therefore may be able to cancel its potential benefi-cial effect on glycemic control based on the effect gas-trin. This may also be one of the reasons why the effects on glycemic control of the two therapies were equivalent.

In this study, significant reduction in TC, LDL-C, and HDL-C was found in combination therapy. It is difficult to clearly explain the reason why these results were obtained. However, it is unlike that lansoprazole influenced the decrease of these lipids levels, because it is reported that PPI therapy for 3 months signifi-cantly increased (but not decreased) TC and LDL-C levels with no change of HDL-C in patients with type 2 diabetes [25]. On the other hand, it is reported that alogliptin can reduce LDL-C levels [26-28], although these levels did not change in some studies [29, 30]. Recently, Ayaori et al. reported that alogliptin sig-nificantly reduced TC, LDL-C, and HDL-C (but not TG), and that it significantly deceased HDL-C, even when compared with sitagliptin [28]. Thus, we spec-ulate that decrease of TC, LDL-C, and HDL-C which were observed in this study was probably associated with the effect of alogliptin, although the precise mech-anism is unknown. In fact, also in alogliptin group, the tendency of decrease of these lipids was found; this


1. Rooman I, Lardon J, Bouwens L (2002) Gastrin stim-ulates β-cell neogenesis and increases islet mass from transdifferentiated but not from normal exocrine pan-creas tissue. Diabetes 51: 686-690.

2. Meier JJ, Butler AE, Galasso R, Rizza RA, Butler PC (2006) Increased islet beta cell replication adjacent to intrapancreatic gastrinomas in humans. Diabetologia 49: 2689-2696.

3. Sheen E, Triadafilopoulos G (2011) Adverse effects of long-term proton pump inhibitor therapy. Dig Dis Sci 56: 931-950.

4. Ligumsky M, Lysy J, Siguencia G, Friedlander Y (2001) Effect of long-term, continuous versus alternate-day omeprazole therapy on serum gastrin in patients treated for reflux esophagitis. J Clin Gastroenterol 33: 32-35.

5. Hu YM, Mei Q, Xu XH, Hu XP, Hu NZ, Xu JM (2006) Pharmacodynamic and kinetic effect of rabeprazole on serum gastrin level in relation to CYP2C19 polymor-phism in Chinese Hans. World J Gastroenterol 12: 4750-4753.

6. Cadiot G, Vissuzanic C, Pospai D, Ruszniewski P, Potet F, Mignon M (1995) Effect of prolonged treatment with proton pump inhibitors on serum gastrin levels and the fundus mucosa. Preliminary results. Gastroenterol Clin Biol 19: 811-817 (In French).

7. Koop H, Klein M, Arnold R (1990) Serum gastrin lev-els during long-term omeprazole treatment. Aliment Phamacol Ther 4: 131-138.

8. Sanders SW, Tolman KG, Greski PA, Jennings DE, Hoyos PA, Page JG (1992) The effects of lansoprazole, a new H+, K(+)-ATPase inhibitor, on gastric pH and serum gastrin. Aliment Phamacol Ther 6: 359-372.

change may have been significant if the study was per-formed in larger number of patients.

Apart from the relative short term-observation, this study has limitations. This study was not dou-ble-blinded using a placebo. In addition, it is known that diabetes occasionally occurs with gastroesoph-ageal reflux disease (GERD). We could not investi-gate the proportion of patients with GERD in this study. Because PPI largely improves GERD clinical symptoms, it may be possible that the appetite of the patients with GERD was improved, resulting in poten-tially worsening glycemic control. This may also explain partially the lack of a synergy effect with PPI and DPP4-inhibitor on glycemic control in this study. Finally, we missed measuring plasma GLP-1 and gas-tric inhibitory polypeptide (GIP) levels in this study. It may have been important to investigate the degrees

of elevation of plasma GLP-1 and GIP levels by these therapies. We admit that this is also one of the limita-tions of this study.

In conclusions, in the current study, we showed that the effect on glycemic control of combination ther-apy with alogliptin and lansoprazole is equal to that of alogliptin mono-therapy in a 3-month study. Longer observation for the effect of combination therapy by DDP4-inhibitor and PPI, and a study using GLP-1 receptor agonist instead of DPP4-inhibitor with PPI would be warranted.

Disclosure Summary

The authors have nothing to disclose. There is no conflict of interest that could be perceived as prejudic-ing the impartiality of the research reported.

References 9. Bödvarsdóttir TB, Hove KD, Gotfredsen CF, Pridal L,

Vaag A, et al. (2010) Treatment with a proton pump inhibitor improves glycaemic control in Psammomys obesus, a model of type 2 diabetes. Diabetologia 53: 2220-2223.

10. Hove KD, Færch K, Bödvarsdóttir TB, Karlsen AE, Petersen JS, Vaag A (2010) Treatment with a proton pump inhibitor improves glycaemic control in type 2 diabetic patients-a retrospective analysis. Diabetes Res Clin Pract 90: e72-e74.

11. Crouch MA, Mefford IN, Wade EU (2012) Proton pump inhibitor therapy associated with lower glycosyl-ated hemoglobin levels in type 2 diabetes. J Am Board Fam Med 25: 50-54.

12. Boj-Carceller D, Bocos-Terraz P, Moreno-Vernis M, Sanz-Paris A, Trincado-Aznar P, Albero-Gamboa R (2011) Are proton pump inhibitors a new antidiabetic drug? A cross sectional study. World J Diabetes 2: 217-220.

13. Singh PK, Hota D, Dutta P, Sachdeva N, Chakrabarti A, et al. (2012) Pantoprazole improves glycemic con-trol in type 2 diabetes: a randomized, double-blind, placebo-controlled trial. J Clin Endocrinol Metab 97: E2105-E2108.

14. Hove KD, Brøns C, Færch K, Lund SS, Petersen JS, et al. (2013) Effects of 12 weeks’ treatment with a proton pump inhibitor on insulin secretion, glucose metabo-lism and markers of cardiovascular risk in patients with type 2 diabetes: a randomized double-blind prospective placebo-controlled study. Diabetologia 56: 22-30.

15. Sadry SA, Drucker DJ (2013) Emerging combinational hormone therapies for the treatment of obesity and


T2DM. Nat Rev Endocrinol 9: 425-433.16. Suarez-Pinzon WL, Power RF, Yan Y, Wasserfall C,

Atkinson M, Rabinovitch A (2008) Combination ther-apy with glucagon-like peptide-1 and gastrin restores normoglycemia in diabetic NOD mice. Diabetes 57: 3281-3288.

17. Suarez-Pinzon WL, Cembrowski GS, Rabinovitch A (2009) Combination therapy with a dipeptidyl peptidase-4 inhibitor and a proton pump inhibitor restores normoglycaemia in non-obese diabetic mice. Diabetologia 52: 1680-1682.

18. Patel V, Joharapurkar A, Gandhi T, Patel K, Dhanesha N, et al. (2013) Omeprazole improves the anti-obesity and antidiabetic effects of exendin-4 in db/db mice. J Diabetes 5: 163-71.

19. Fosgerau K, Jessen L, Lind Tolborg J, Østerlund T, Schæffer Larsen K, et al. (2013) The novel GLP-1-gastrin dual agonist, ZP3022, increases β-cell mass and prevents diabetes in db/db mice. Diabetes Obes Metab 15: 62-71.

20. Seino Y, Fujita T, Hiroi S, Hirayama M, Kaku K (2011) Efficacy and safety of alogliptin in Japanese patients with type 2 diabetes mellitus: a randomized, double blind, dose-ranging comparison with placebo, followed by a long-term extension study. Curr Med Res Opin 27: 1781-1792.

21. Treatment guide for diabetes edited by Japan Diabetes Society 2007 (Bunkodo Co., Ltd, Tokyo, Japan).

22. Singh P, Indaram A, Greenberg R, Visvalingam V, Bank K (2000) Long term omeprazole therapy for reflux esophagitis: follow-up in serum gastrin levels, EC cell hyperplasia. World J Gastroentrol 6: 789-792.

23. Brunner G, Athmann C, Schneider A (2012) Long-term open-label trial: safety and efficacy of continous main-tenance treatment with pantprazole for up to 15 years in severe acid-peptic disease. Aliment Pharmacol Ther 36: 37-47.

24. Wallace JL, Syer S, Denou E, de Palma G, Vong L, et

al. (2011) Proton pump inhibitors exacerbate NSAID-induced small intestinal injury by inducing dysbiosis. Gastroenterology 141: 1314-1322.

25. Hove KD, Brøns C, Færch K, Lund SS, Petersen JS, et al. (2013) Effects of 12 weeks’ treatment with a proton pump inhibitor on insulin secretion, glucose metabo-lism and markers of cardiovascular risk in patients with type 2 diabetes: a randomized double-blind prospective placebo-controlled study. Diabetologia 56: 22-30.

26. Kaku K, Itayasu T, Hiroi S, Hiryama M, Seino Y (2011) Efficacy and safety of alogliptin added to pioglitazone in Japanese patients with type 2 diabetes: a randomized, double-blind, placebo-controlled trial with an open-label long-term extension study. Diabetes Obes Metab 13: 1028-1035.

27. Seino Y, Fujita T, Hiroi S, Hirayama M, Kaku K (2011) Alogliptin plus voglibose in Japanese patients with type 2 diabetes: a randomized, double blind, placebo-con-trolled trial with an open-label, long-term extension . Curr Med Res Opin 27: Supple 3: 21-29.

28. Ayaori M, Iwakami N, Uto-Kondo H, Sato H, Sasaki M, et al. (2013) Dipeptidyl peptidase-4 inhibitors attenuate endothelial fuction as evaluated by flow-mediated vaso-dilatation in type 2 diabetic patients. J Am Heart Assoc 2: e003277.

29. Kaku K, Itayasu I, Hiroi S, Hirayama M, Seino Y (2011) Efficacy and safety of alogliptin added to pioglitazone in Japanese patients with type 2 diabetes: a randomized, double-blind, placebo-controlled trial with an open-label long-term extension study. Diabetes Obes Metab 13: 1028-1035.

30. Seino Y, Miyata Y, Hiroi S, Hirayama M, Kaku K (2012) Efficacy and safety of alogliptin added to met-formin in Japanese patients with type 2 diabetes: a ran-domized, double-blind, placebo-controlled trial with an open-label, long-term extension study. Diabetes Obes Metab 14: 927-936.

Effect of combination therapy with alogliptin and ...

Documents

Transcript of Effect of combination therapy with alogliptin and ...